DHL Supply Chain Deploys Agentic AI to Automate Calls, Email, and Scheduling
DHL Supply Chain has partnered with HappyRobot to roll out agentic AI that handles phone calls, email, and routine communications with partners and internal teams. The goal is simple: speed up coordination, remove busywork, and let people focus on exceptions and higher-value work.
These AI agents act independently to solve specific tasks. One voice agent can schedule drivers on the phone while another manages shipment email threads end-to-end.
Where It's Live and What It's Doing
DHL has already deployed agents across several markets. Current use cases include appointment scheduling, driver follow-up calls, and high-priority warehouse coordination.
The company expects to save millions of minutes on the phone annually and handle hundreds of thousands of emails with agents. That's meaningful capacity back to the floor without adding headcount.
How Leadership Frames It
"Building on our extensive operational experience with data analytics, robotic process automation, and self-learning software tools, we are now integrating AI agents to drive greater process efficiency for customers while making operational roles more engaging and rewarding for employees by automating repetitive and time-consuming tasks such as manual data entry, routine scheduling and standardized communications," said Sally Miller, chief information officer at DHL Supply Chain.
People Impact: Less Busywork, Better Retention
There's public concern about AI displacing jobs-more than seven in 10 Americans are worried about that, according to a Reuters/Ipsos poll. DHL's position is that agents remove repetitive tasks and help teams focus on work that needs human judgment.
"At DHL Supply Chain, our people are at the heart of everything we do," said Lindsay Bridges, executive vice president of human resources. "AI agents help us relieve our teams from repetitive, time-consuming tasks and give them space to focus on meaningful, high-value work... these technologies allow us to maintain-and even improve-responsiveness, customer centricity and service consistency, while making roles more attractive and sustainable."
Signal from the Market
HappyRobot also supports Flexport, Ryder, and Circle Logistics. Flexport's CEO, Ryan Petersen, has discussed using voice agents to check shipment status and timing-agents that gather details from providers and relay information to humans. He expects a near future where agents coordinate directly with each other over the phone.
Pablo Palafox, CEO of HappyRobot, summed up the direction: "At HappyRobot, we envision AI workers coordinating global supply chain operations-not just moving data, but actively managing workflows... DHL recognized early on the potential of AI agents as a new operating layer-one that brings speed, visibility and consistency to logistics."
What Operations Leaders Should Do Next
- Pick narrow, high-volume use cases: dock scheduling, check calls, ASN/PO status emails, detention alerts.
- Define guardrails: escalation rules, allowed systems, approved scripts, compliance boundaries.
- Start with a shadow period: AI drafts or conducts calls under supervision; humans review before sending or confirm outcomes post-call.
- Instrument everything: track handle time, first-contact resolution, SLA adherence, error rate, exception rate, and employee time saved.
- Integrate with your stack: TMS, WMS, email, and telephony. Keep a unified activity log tied to orders/loads.
- Close the loop: feed incorrect outcomes back for tuning; update playbooks and prompts weekly during pilot.
Risk and Controls to Put in Place
- Compliance: ensure consent for call recording, data retention, and PII handling. Align with your legal guidance and policy.
- Quality: require human-in-the-loop for new scenarios; graduate to autonomy only after hitting quality thresholds.
- Transparency: clear handoff to a human when confidence is low or a policy boundary is hit.
- Auditability: keep transcripts, message logs, and action trails linked to orders for accountability and training.
- Safety baselines: adopt a simple risk framework (e.g., NIST AI RMF) for use-case approval and ongoing review. See NIST AI RMF.
KPI Snapshot for Agentic AI in Ops
- Call minutes saved per week and per site
- Emails auto-resolved vs. escalated
- Appointment lead time reduced and no-show rate
- First-contact resolution and average handle time
- Exception rate by category and time-to-resolution
- Employee NPS/retention in roles affected by automation
- Customer response time and SLA adherence
Practical Rollout Plan (90 Days)
- Weeks 1-2: Select one lane/site and one use case. Map scripts, data fields, and escalation paths.
- Weeks 3-6: Run supervised mode. Measure baseline vs. assisted numbers. Tighten prompts and workflows.
- Weeks 7-10: Gradual autonomy with clear quality gates. Add a second use case only after stability.
- Weeks 11-13: Document SOPs, training, and audit steps. Prep for multi-site replication.
If You're Building Team Capability
Upskill planners, coordinators, and supervisors on AI-assisted workflows and prompt standards. A small investment in training shortens tuning cycles and improves results.
For role-based learning paths on AI in operations, see Complete AI Training: Courses by Job.
Your membership also unlocks:
